Open WebUI is a self-hosted interface for running LLMs locally or via API. Any model works, but it benefits from vision support, streaming, and large context for document analysis.
Best Models for Open WebUI
Top 15 by tool-optimized score
Scored by: benchmark performance (90%) from MMLU, GPQA, HumanEval, SWE-bench, and 15+ standardized evaluations, with capabilities and context as tiebreakers (10%).
| # | Model | Score | Output $/M |
|---|---|---|---|
| 1 | Gemma 4 31B Arena Elo: 1451 | 93 | $0.350 |
| 2 | Gemma 4 26B A4B Arena Elo: 1438 | 92 | $0.330 |
| 3 | MiMo-V2.5 Arena Elo: 1433 | 91 | $0.280 |
| 4 | Kimi K2.6 Arena Elo: 1460 | 91 | $3.41 |
| 5 | MiniMax M3 Arena Elo: 1448 | 90 | $1.20 |
| 6 | Grok 4.3 Arena Elo: 1446 | 90 | $2.50 |
| 7 | Gemma 4 31B (free) | 90 | Free |
| 8 | Qwen3 VL 235B A22B Instruct Arena Elo: 1415 | 90 | $0.880 |
| 9 | Gemini 3.5 Flash Arena Elo: 1477 | 89 | $9.00 |
| 10 | Qwen3.6 Plus Arena Elo: 1444 | 89 | $1.95 |
| 11 | Qwen3.5 397B A17B Arena Elo: 1444 | 89 | $2.45 |
| 12 | Grok 4.20 Multi-Agent | 88 | $2.50 |
| 13 | Grok 4.20 | 88 | $2.50 |
| 14 | Gemini 3.1 Flash Lite Preview Arena Elo: 1432 | 88 | $1.50 |
| 15 | Qwen3.5-Flash Arena Elo: 1397 | 88 | $0.260 |
| 16 | GPT-5.1-Codex-Mini | 88 | $2.00 |
| 17 | Qwen3.5-122B-A10B Arena Elo: 1417 | 87 | $2.08 |
| 18 | Gemini 3.1 Pro Preview Custom Tools | 87 | $12.00 |
| 19 | GPT-4o-mini HumanEval: 87.2% | 87 | $0.600 |
| 20 | Claude Fable 5 SWE-bench: 95% | 86 | $50.00 |
| 21 | Gemma 4 26B A4B (free) | 86 | Free |
| 22 | Qwen3.5-27B Arena Elo: 1409 | 86 | $1.56 |
| 23 | GPT-5.2-Codex | 86 | $14.00 |
| 24 | Gemini 2.5 Flash Lite Preview 09-2025 | 86 | $0.400 |
| 25 | Gemini 2.5 Flash Lite | 86 | $0.400 |
| 26 | Claude Opus 4.8 (Fast) | 85 | $50.00 |
| 27 | Claude Opus 4.7 (Fast) | 85 | $150.00 |
| 28 | Mistral Medium 3.5 Arena Elo: 1426 | 85 | $7.50 |
| 29 | Qwen3.5-35B-A3B Arena Elo: 1396 | 85 | $1.00 |
| 30 | GPT-5.3-Codex | 85 | $14.00 |
| 31 | GPT-5.2 Chat Arena Elo: 1475 | 85 | $14.00 |
| 32 | GPT-5.1-Codex-Max | 85 | $10.00 |
| 33 | GPT-5.1-Codex | 85 | $10.00 |
| 34 | GPT-5 Codex | 85 | $10.00 |
| 35 | GPT-5.4 Pro | 84 | $180.00 |
| 36 | Claude Opus 4.1 Arena Elo: 1449 | 84 | $75.00 |
| 37 | Claude Opus 4.6 (Fast) | 83 | $150.00 |
| 38 | GPT-5.4 Nano | 83 | $1.25 |
| 39 | GPT-5.4 Mini | 83 | $4.50 |
| 40 | Gemini 3 Flash Preview SWE-bench: 78% | 83 | $3.00 |
| 41 | GPT-5.2 Pro | 83 | $168.00 |
| 42 | GLM 4.6V Arena Elo: 1377 | 83 | $0.900 |
| 43 | Gemini 2.5 Pro Preview 06-05 | 83 | $10.00 |
| 44 | Gemini 2.5 Pro Preview 05-06 | 83 | $10.00 |
| 45 | Claude Opus 4.8 SWE-bench: 88.6% | 82 | $25.00 |
| 46 | GPT-5.5 Pro SWE-bench: 88.7% | 82 | $180.00 |
| 47 | GPT-5.5 SWE-bench: 88.7% | 82 | $30.00 |
| 48 | MiMo-V2.5-Pro Arena Elo: 1466 | 82 | $0.870 |
| 49 | Claude Opus 4.7 SWE-bench: 87.6% | 82 | $25.00 |
| 50 | o4 Mini Deep Research | 82 | $8.00 |
Based on our analysis of coding benchmarks, capability matching, and pricing, Gemma 4 31B currently ranks #1 for Open WebUI. Rankings are rebuilt as benchmark, pricing, and provider data refresh.
We score models using benchmark performance (90%) from LMArena, HumanEval, SWE-bench, MMLU, and 15+ standardized evaluations. Capabilities and context serve as tiebreakers (10%). Only models with the capabilities Open WebUI needs are included in the tool-specific rankings.
We currently track 316 AI models compatible with Open WebUI. This includes models from OpenAI, Anthropic, Google, DeepSeek, and other providers accessible via API.
Many open-source models are compatible with Open WebUI through API providers like OpenRouter, Together AI, and Groq. Check our rankings to see which open-source models perform best.
Rankings refresh whenever the underlying benchmark, pricing, and catalog sources refresh. That means some signals update faster than others, and the page reflects the latest verified source data available.